flies and Hooks. 99 



feather of the wing, or the sheen of the silver bodies 

 could be seen. When a dark-colored body was visible 

 the hook was invariably at least equally conspicuous. 

 Silver tinsel appeared to be more efficient than gold, as 

 might be expected from its higher reflecting power, 

 though both held their own well. 



This summarizes such results of these experiments as 

 seem to me to be of general interest, and which I am 

 willing to characterize as facts. 



Applying them to the many and apparently irrecon- 

 cilable differences of opinion in regard to the varieties of 

 flies preferred by salmon, and bearing in mind my belief 

 that salmon take the fly as and for food, it seems to me 

 an explanation may be found which goes far toward ac- 

 counting for many of these apparent discrepancies. 



When would we naturally expect the salmon to take 

 the fly ? Clearly when, though visible, it resembles as 

 little as possible what it really is a flower of fancy and 

 no production of nature. W T e find these conditions filled 

 when the size or colors, or both, of the fly are so adjusted 

 to the depth and transparency of the water and the light 

 which falls upon it, that the details of the fly are obscure, 

 and only the general effect of a living and moving object 

 is produced. Should the attention of a salmon be then 

 directed to the fly, how can it do otherwise than attribute 

 the effect produced to that living object, with which it is 

 familiar, which most closely resembles it ? And should 

 that familiar object be one which appeals to its appetite 

 at the time, the salmon would then naturally, as it seems 

 to me, make an effort to take it. 



On the other hand, should the fly be so large, or its 



